forked from gmittal/gmittal.github.io
-
Notifications
You must be signed in to change notification settings - Fork 0
/
index_b.html
243 lines (223 loc) · 11.7 KB
/
index_b.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
<!doctype html>
<html lang="en">
<head>
<title>Blake Bullwinkel</title>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="icon" type="image/png" href="/static/marble.png">
<link rel="stylesheet" type="text/css" href="styles.css">
</head>
<body>
<div class="profile">
<div class="profile-pic">
<img src="static/me.JPG" />
</div>
<div class="profile-text">
<h1>Blake Bullwinkel</h1>
<h3>Data Scientist</h3>
<div class="social">
<a href="mailto:[email protected]" rel="noreferrer">
<img src="static/email.svg" id="in" alt="Email" />
</a>
<a href="https://github.com/blakebullwinkel" target="_blank">
<img src="static/github.svg" id="gh" alt="GitHub" />
</a>
<a href="https://www.linkedin.com/in/blakebullwinkel/" target="_blank">
<img src="static/linkedin.png" id="in" alt="LinkedIn" />
</a>
<a href="https://scholar.google.com/citations?user=EwwZ0JMAAAAJ&hl=en&oi=sra" target="_blank">
<img src="static/scholar.svg" id="in" alt="Scholar" />
</a>
<a href="https://medium.com/@blakebullwinkel" target="_blank">
<img src="static/medium.png" id="me" alt="Medium" />
</a>
</div>
</div>
</div>
<p>
I'm a Data Scientist at Microsoft and recently completed a master's at Harvard's
<a href="https://iacs.seas.harvard.edu/" target="_blank">IACS</a>, where I was advised by
<a href="https://iacs.seas.harvard.edu/people/pavlos-protopapas" target="_blank">Pavlos Protopapas</a>.
My <a href="static/thesis.pdf" target="_blank">thesis</a> focused on using GANs to solve differential
equations. In general, I'm interested in applied machine learning research, particularly when it overlaps
with human-centered domains. I spend my free time running, practicing piano, and working on side projects.
</p>
<p>
In the spring of 2020, I earned my BA from Williams College, where I double majored in math and Chinese.
I spent my junior year studying at the University of Oxford, and before my undergrad years lived in
Hong Kong, Singapore, and Sydney. You can find my <a href="static/resume.pdf" target="_blank">resume</a>
here and should feel free to reach out!
</p>
<h2>Research</h2>
<p>
<span class="pub">Transfer Learning with Physics-Informed Neural Networks for Efficient Simulation of Branched Flows</span><br />
Raphael Pellegrin, Blake Bullwinkel, Marios Mattheakis, Pavlos Protopapas<br />
NeurIPS Workshop on Machine Learning and the Physical Sciences, 2022<br />
<a class="button" href="https://arxiv.org/abs/2211.00214" target="_blank">arXiv</a>
<a class="button" href="https://neurips.cc/media/PosterPDFs/NeurIPS%202022/56854.png" target="_blank">Poster</a>
</p>
<p>
<span class="pub">DEQGAN: Learning the Loss Function for PINNs with Generative Adversarial Networks</span><br />
Blake Bullwinkel, Dylan Randle, Pavlos Protopapas, David Sondak<br />
ICML Workshop on AI for Science (AI4Science), 2022<br />
<a class="button" href="https://arxiv.org/abs/2209.07081" target="_blank">arXiv</a>
<a class="button" href="static/deqgan_poster_ai4science.pdf" target="_blank">Poster</a>
</p>
<p>
<span class="pub">Evaluating the Fairness Impact of Differentially Private Synthetic Data</span><br />
Blake Bullwinkel, Kristen Grabarz, Lily Ke, Scarlett Gong, Chris Tanner, Joshua Allen<br />
ICML Workshop on Theory and Practice of Differential Privacy (TPDP), 2022<br />
<a class="button" href="https://arxiv.org/abs/2205.04321" target="_blank">arXiv</a>
<a class="button" href="static/dp_fairness_poster.pdf" target="_blank">Poster</a>
</p>
<hr style="height:4px; visibility:hidden;" />
<h2>Projects</h2>
<div class="project-container">
<p>
<span class="pub">Marble Groceries</span><br />
Led a team of software engineers and business development managers to develop an iOS app that helps users understand the environmental impact of their grocery purchases by scanning product barcodes.<br />
<a class="button" href="https://apps.apple.com/us/app/marble-groceries/id1606374023?platform=iphone" target="_blank">App Store</a>
</p>
<div class="project-img">
<img src="static/marble-thumbnail.png" />
</div>
</div>
<div class="project-container">
<p>
<span class="pub">Linearized Neural Nets for Transfer Learning with GPs</span><br />
Implemented the method for transfer learning proposed in <i>Fast Adaptation with Linearized Neural Networks</i>, a <a href="https://arxiv.org/pdf/2103.01439.pdf" target="_blank">2021 paper</a> by Maddox et al., in TensorFlow and performed experiments to test its practical utility.<br />
<a class="button" href="static/linearized-nets-notebook.html" target="_blank">Notebook</a>
<a class="button" href="https://github.com/blakebullwinkel/am207_final_project" target="_blank">Code</a>
</p>
<div class="project-img">
<img src="static/gp-plot.png" />
</div>
</div>
<div class="project-container">
<p>
<span class="pub">Classifying the Sounds of NYC</span><br />
Trained and tuned a variety of models to classify audio clips recorded around New York City from the UrbanSound8k dataset into ten different classes.<br />
<a class="button" href="static/ml-report.pdf" target="_blank">Report</a>
<a class="button" href="static/ml-notebook.html" target="_blank">Notebook</a>
</p>
<div class="project-img">
<img src="static/mel-plot.png" />
</div>
</div>
<div class="project-container">
<p>
<span class="pub">Modeling ASA Section Membership</span><br />
Constructed binary response generalized linear models to predict whether or not members of the American Statistical Association belonged to at least one section.<br />
<a class="button" href="static/glm-report.pdf" target="_blank">Report</a>
<a class="button" href="https://github.com/yingchenliu98/stat149_final_project" target="_blank">Code</a>
</p>
<div class="project-img">
<img src="static/gam-smooths.png" />
</div>
</div>
<div class="project-container">
<p>
<span class="pub">Forecasting and Classifying Mice Microbiomes</span><br />
Used deep learning to forecast qPCR time series and classify mouse microbiomes into healthy and infected groups based on data provided by researchers at Brigham and Women's Hospital.<br />
<a class="button" href="static/microbiome-notebook.html" target="_blank">Notebook</a>
</p>
<div class="project-img">
<img src="static/qpcr.PNG" />
</div>
</div>
<div class="project-container">
<p>
<span class="pub">Woof Woof! Computer Vision & NLP App for Austin Pets Alive</span><br />
Attended workshops at the 2021 <a href="https://www.computefest.seas.harvard.edu/" target="_blank">IACS ComputeFest</a> to build a web app that allows users to "chat" with and search for visually similar dogs in the Austin Pets Alive animal shelter by leveraging backend NLP and computer vision models.<br />
<a class="button" href="https://github.com/blakebullwinkel/woof-woof-app" target="_blank">Code</a>
</p>
<div class="project-img">
<img src="static/woofwoof.PNG" />
</div>
</div>
<div class="project-container">
<p>
<span class="pub">Wildfire Risk Prediction & Response Optimization</span><br />
Trained tree-based classification models on historical wildfire and weather data to predict the fire risk for a given county and month in California and used mixed-integer programming to determine the optimal assignment of limited firefighters across the state, based on total cost.<br />
<a class="button" href="static/ai-report.pdf" target="_blank">Report</a>
<a class="button" href="https://github.com/teresadatta/CA-Wildfire-Risk-Prediction-and-Optimization" target="_blank">Code</a>
</p>
<div class="project-img">
<img src="static/wildfire-map.png" />
</div>
</div>
<div class="project-container">
<p>
<span class="pub">Predicting the Outcome of the 2020 Election</span><br />
Built k-NN and regularized logistic regression models to predict the outcomes of the 2020 presidential and congressional elections using historical, polling, and fundamentals data.<br />
<a class="button" href="static/election-report.pdf" target="_blank">Report</a>
<a class="button" href="https://github.com/blakebullwinkel/2020-Election-Prediction" target="_blank">Code</a>
</p>
<div class="project-img">
<img src="static/election-map.png" />
</div>
</div>
<div class="project-container">
<p>
<span class="pub">DreamDiff Python Package</span><br />
Worked in a team of three to develop a Python package that implements forward-mode automatic differentiation (AD), root-finding, optimization, and quadratic spline interpolation.<br />
<a class="button" href="https://github.com/autodiffdreamteam/cs107-FinalProject" target="_blank">Repo</a>
<a class="button" href="https://pypi.org/project/DreamDiff/1.0.1/" target="_blank">PyPI</a>
</p>
<div class="project-img">
<img src="static/grad-desc.gif" />
</div>
</div>
<div class="project-container">
<p>
<span class="pub">Analysis of Wildfires, Air Quality, and Public Health</span><br />
Conducted time series analysis in R to link spikes in PM2.5 concentration to specific wildfire events in California and used major axis regression to explore correlations between air quality and public health outcomes.<br />
<a class="button" href="static/eps-slides.pdf" target="_blank">Slides</a>
<a class="button" href="https://github.com/Camerajonas/EPS236-Wildfire-Project" target="_blank">Code</a>
</p>
<div class="project-img">
<img src="static/pm-season.png" />
</div>
</div>
<div class="project-container">
<p>
<span class="pub">Predicting Agricultural Crop Quality</span><br />
Built a variety of regression models to predict the quality of crops based on agricultural and weather data.<br />
<a class="button" href="static/pepsico-slides.pdf" target="_blank">Slides</a>
<a class="button" href="https://github.com/blakebullwinkel/pepsico-data-science" target="_blank">Code</a>
</p>
<div class="project-img">
<img src="static/yeo-johnson.png" />
</div>
</div>
<div class="project-container">
<p>
<span class="pub">Modeling Electricity Consumption in the US</span><br />
Built linear regression models to predict household electricity consumption in the US from various residential characteristics.<br />
<a class="button" href="static/electricity-report.pdf" target="_blank">Report</a>
<a class="button" href="https://github.com/blakebullwinkel/electricity-consumption" target="_blank">Code</a>
</p>
<div class="project-img">
<img src="static/electricity-histograms.png" />
</div>
</div>
<div class="project-container">
<p>
<span class="pub">Early Epidemiological Model Parameters for COVID-19</span><br />
Modeled early-stage COVID-19 case data in mainland China using systems of ordinary differential equations.<br />
<a class="button" href="static/covid-slides.pdf" target="_blank">Slides</a>
<a class="button" href="https://github.com/blakebullwinkel/covid-modeling" target="_blank">Code</a>
</p>
<div class="project-img">
<img src="static/covid-histograms.png" />
</div>
</div>
<br />
<footer>
<p style="text-align:right;font-size:small;">
Website template by <a href="https://mittal.ai">Gautam Mittal</a><br>
Last updated November 2022
</p>
</footer>
</body>
</html>